The relationship between magnetism, magnetics, and microstructure is discussed from the viewpoint of three different theoretical techniques. A dynamic micromagnetic theory is used to show that hysteresis can be accurately predicted from structural properties without the need for adjustable parameters. Analysis of these results shows that microstructure can be the dominant contributor to micromagnetic properties. Electronic structure calculations including the spin-orbit interaction show that the predicted value of the magnetic anisotropy in Co/Pd superlattices is only obtained in samples grown under ideal conditions; other samples possess less anisotropy probably owing to a combination of misorientation, poor interfaces and voids. Finally, surfaces are shown to lower magnetization in a super-exchange material (owing to thermal fluctuations) but raise it in a polarizable transition metal.